/srv/irclogs.ubuntu.com/2020/05/28/#launchpad-dev.txt

tomwardilluse node-sass for (nearly) no-op CSS compilation: https://code.launchpad.net/~twom/launchpad/+git/launchpad/+merge/38470609:32
tomwardillthe output file isn't exactly the same as the current one, but it contains all the same rules afaics (and the site seems to work...)09:33
tomwardillthe watch command has some slightly strange output due to mishandling of subdirectories and imports, but it works and I think just needs some layout tweaks that I'll hopefully get to next09:33
cjwatsontomwardill: Can we remove lib/lp/scripts/utilities/js/combo.py and lib/lp/scripts/utilities/js/tests/test_combo.py as well?09:44
cjwatsonI guess all of lib/lp/scripts/utilities/js/tests/ in fact09:44
tomwardillcjwatson: I think that's still used for the JS files09:44
cjwatsonAre you sure?  It was at one point, but grep doesn't find any other references09:45
cjwatsonUnless I'm misreading09:45
cjwatsonWorth a bit of hunting around to see what else we can remove, anyway09:45
tomwardillI'll have another poke at it09:46
* tomwardill randomly deletes things to see what breaks09:46
lifelesssomewhat randomly, is LP using pybars? There are forks of it adding python3 and newer handlebars.js features10:18
lifelessI don't know if anyone reached out to the LP team about it, but it might be an idea to hand over the pypi project to one of the forks if not10:19
* lifeless hands over whatever blessing might possibly be needed for that10:19
cjwatsonlifeless: no10:19
cjwatsonat least I don't think so10:20
cjwatsonwe use pystache in the buglisting tests10:21
cjwatsonand a copy of mustache.js I think10:22
tomwardillcjwatson: removed those files: https://code.launchpad.net/~twom/launchpad/+git/launchpad/+merge/384706 dont' seem to be able to remove any further without a re-engineering of the jsbuild stuff10:27
tomwardillI'd like CSSComboFile to go away, but stuff breaks if I try10:27
cjwatsonShould hopefully be possible to redo Builder.build_skin etc. using scss, but doesn't have to be in this MP10:31
tomwardillyeah, that's where I'd like to get it eventually10:32
tewardcjwatson: just as an FYI for the bug i filed about code review DMARC compliance, I'm still working on it, but some more major IT Security related things came up @ employers so i had to address that stuff heh13:27
tewardin case anyone was asking :)13:27
teward(i'm still working on that branch heh)13:28
cjwatsonNobody has been asking, but thanks :)13:28
cjwatsonNo immediate rush, whenever you have time13:28
tewardyeh the status change on the bug is what caught my attention :)13:34
teward(since i filed it xD)13:34
tewardcjwatson: i'm not subscribed to the bugs and questions list, has there been any flak from the Community At Large with the email change yet?  Or edge cases and bugs and stuff?13:37
cjwatsonteward: not a word13:37
tewardlucky us.13:37
tewardguess the vast majority of people don't care enough to keep an eye on it as long as it works heh13:37
cjwatsonOh, one blog comment13:38
cjwatsonBasically asking for name mangling13:38
tewardname mangling how?  (Comments are invisible heh)13:39
cjwatsonHm?  No they're not, https://blog.launchpad.net/notifications/bug-emails-now-use-the-bugs-address-in-the-from-header13:40
tewardthey are if you ain't logged in13:41
cjwatsonI'm not logged in13:41
* teward just logged in with SSO and can see it now13:41
cjwatsonI checked in w3m13:41
tewardhmm13:41
tewardwonder if my system is being weird.13:41
teward*goes to check it*13:41
cjwatsonshift-reload13:41
cjwatsonSame sort of thing we discussed as maybe a possibility, anyway.  I don't think it's immediately urgent but we should probably have a bug report for it13:41
tewardi mean13:42
tewardthat's pretty fixable13:42
cjwatsonAnd TBH I'd rather avoid name mangling if we can13:42
tewardbut that would require the level of complexity we had before13:42
tewardwith special case exclusions, etc.13:43
cjwatsonThe fact that GitHub doesn't do it means we're not alone13:43
tewardyeah most places dno't13:43
tewarddon't*13:43
tewardthose that DO have mangling in line do that as a design decision in their systems / lists13:43
cjwatsonAnyway, we don't need to discuss it in depth here IMO13:43
tewardyep13:44
teward*goes to consume breakfast*13:44
tomwardillhuh, css to scss converters exist14:46
tomwardillhow badly can that go?14:46
SpecialK|Canon<cthulhu/>14:46
tomwardillweirdly, it's... not that bad?14:47
pappacenatomwardill: I don't know... maybe it would change randomly the color of dropdowns? 👀14:47
tomwardillah, from 'grey' to 'slightly not that grey'?14:47
* pappacena hahaha!14:47
SpecialK|Canon;P14:50
tomwardillI dropped 'base.css' into https://css2sass.herokuapp.com/14:50
tomwardilland the result is actually readable14:50
tomwardill(at least to me, as someone who wouldn't know good scss if it wrote me a letter)14:51
tomwardillhmm, actually going to go with that15:00
SpecialK|CanonOne day I'll ship a tool like that and have it embed zalgo in 0.01% of responses or sth, see who notices ;)15:01
* tomwardill pays careful attention to any SpecialK|Canon MP15:02
SpecialK|Canonlol15:02
tomwardillwow, we have some fun about overriding link colours15:21
tomwardill"ImportError: No module named PIL"16:04
tomwardill...not an error I was expecting in this work, I must admit16:04
cjwatsonThat's linked using link-system-packages.py16:07
cjwatsonHave you done 'make compile' in that tree recently?16:07
tomwardillyeah, think I messed up my env16:17
tomwardillmake clean fixed it16:17
tomwardillfound a bug, the combo code rewrites relative urls, so removing it... doesn't16:17
tomwardillso css declared icons now 40416:18

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!